Adaptive Two-level Task Assignment for Human-agent Teams in Dynamic and Adversarial Environments

The goal of this project is the development of an adaptive and decentralized two-level task assignment (TLTA) framework for heterogeneous human-agent teams (HAT) in dynamic and uncertain combat environments. In the proposed TLTA framework, the members of a HAT as well as the tasks are classified into high (coarse/abstract) level and low (fine/physical) level, depending on the types, capabilities, and traits of the humans and agents. The proposed high-level and low-level components of the TLTA algorithm will utilize game-theoretic techniques equipped with auxiliary optimization and discrete path planning methods in order to address hierarchical task assignment problems subject to basic task constraints.